Specific nutrition evaluation of patients with advanced chronic liver disease: Development and validation.

Abstract

Malnutrition is frequently identified in patients with advanced chronic liver disease (ACLD), and its early identification is necessary for effective nutrition treatment. The aim of this study was to develop and validate a tool for specific nutrition evaluation of patients with ACLD (SNE-ACLD). SNE-ACLD was developed by consensus among experts using Delphi technique. The initial proposal for the SNE-ACLD had six domains (history of weight loss, changes in food intake, gastrointestinal symptoms, changes in functional capacity, presence of complications of liver disease, and a nutrition-focused physical examination) and 11 items. Fifteen experts participated in content validation. In a cross-sectional study design, the new tool was applied to 129 inpatients and outpatients from a gastrohepatology unit. Nutrition status was evaluated with SNE-ACLD and subjective global assessment by one researcher. Content validation and semantic analysis were obtained by content validity index. To verify accuracy of SNE-ACLD, the sensitivity, specificity, Youden index, and area under the receiver operating characteristic curve (AUC) were calculated. After five evaluative sequences conducted by experts, experts excluded the domain for history of weight loss and its respective item. The final version of SNE-ACLD consists of five domains and 10 items. The new instrument showed good accuracy in identifying any level of malnutrition (AUC = 0.83; 95% CI, 0.76-0.91) and severe malnutrition (AUC = 0.90; 95% CI, 0.79-1.00). SNE-ACLD can be used in nutrition assessment of patients with liver disease. Future works should investigate its agreement with other methods and its predictive value.